Comparing the effectiveness of a number of statistical methods to equalize the test scores using the common anchor design in the light of some variables

Abstract

The present study aimed to compare the effectiveness of three statistical methods to equalize the test scores which is Levin way for watching signs and Ticker linear way and Unsmoothed Unequipercetile Equating under different experimental conditions whence adjusting the power distribution. The sub-sample has reache ; the small (500), medium (1000), and large (1500), were also conducted at the level of the distribution of natural power and regular.
The study found the following results:
The standard error rates vary depending on the method of equivalent grades. Accordingly, for ways of equation scores vary according to the bilateral interaction between the equation way and capacity distribution. The scores equation vary in the bilateral interaction between the way the equation, and the sample size. Accordingly, the ways equation vary in the triple interaction between the way the equation and distribution capacity and the size of the sample.
Accordingly, the researcher recommends the following:
1. Not shorten in the equation of test scores on the specific way; because of the different ways the effectiveness of the equation depending on the nature of the test models, where there is no way is the best in all situations.
2. Training interested in conducting tests on the use of software that facilitates the process of making the equation between the scores of experimental models.